Improve agentic workflows with enterprise search utilizing Kore.ai and Amazon Q Enterprise

This put up was written with Meghana Chintalapudi and Surabhi Sankhla of Kore.ai.

As organizations wrestle with exponentially rising volumes of knowledge distributed throughout a number of repositories and functions, staff lose vital time—roughly 30% in accordance with the International Data Corporation (IDC)—trying to find info that may very well be spent on higher-value work. The complexity of contemporary enterprise information networks calls for options that may effectively combine, course of, and ship actionable insights throughout disparate methods.

On this put up, we exhibit how organizations can improve their worker productiveness by integrating Kore.ai’s AI for Work platform with Amazon Q Enterprise. We present learn how to configure AI for Work as an information accessor for Amazon Q index for unbiased software program distributors (ISVs), so staff can search enterprise data and execute end-to-end agentic workflows involving search, reasoning, actions, and content material era. We discover the important thing advantages of this integration, together with superior search capabilities throughout greater than 90 enterprise connectors and learn how to lengthen agentic experiences on high of a search basis. The put up features a step-by-step implementation information that will help you arrange this integration in your atmosphere.

Elements of the mixing

Kore.ai is a number one Enterprise AI platform constantly acknowledged by Gartner as a pacesetter in conversational AI. With three key Kore.ai choices, AI for Work, AI for Course of, and AI for Service, enterprises can construct and deploy AI options based mostly on their enterprise wants. The AI for Work platform helps staff be extra productive by making it doable to look throughout functions, take context-aware actions, generate content material, and automate repetitive duties. The platform goes past standalone search to ship complete agentic orchestration and workflows, serving to staff comply with up with purchasers, ship weekly updates, or analysis and write advertising content material with a single command. With AI for Work, your staff can create easy no-code brokers whereas your admins have the pliability to create extra superior low-code or pro-code brokers. AI for Process, however, automates knowledge-intensive enterprise processes end-to-end. AI for Service helps organizations ship differentiated customer support experiences via self-service, proactive outreach campaigns, and agent help.

Amazon Q index for ISVs is a robust, managed vector search service that helps seamless integration of generative AI functions with prospects’ enterprise information via a unified, safe index. ISVs can entry and retrieve related content material via the SearchRelevantContent API for cross-application information retrieval while not having direct entry or particular person indexing of every information supply, whereas prospects retain full management over information entry and governance.

When mixed with extra search connectors provided by AI for Work platform and its skill to create and orchestrate brokers, organizations acquire a whole resolution that transforms how staff entry enterprise information and execute duties end-to-end. The next video reveals one such agentic expertise in motion, the place the AI for Work interface seamlessly orchestrates brokers to assist a gross sales government put together for a shopper assembly—compiling info from Amazon Q index and AI for Work connectors, summarizing speaking factors, and sending them as an e-mail, all from a single question.

Advantages for enterprises

Enterprises usually wrestle with fragmented information entry and repetitive guide duties that decelerate essential enterprise processes. For instance, think about a situation the place a product supervisor must compile quarterly function requests—with the mixing of Kore.ai’s AI for Work and Amazon Q index, they’ll immediately collect requests from Salesforce, help tickets, and JIRA; routinely generate a structured roadmap; and schedule stakeholder conferences, all with a single question. This seamless integration adjustments the way in which enterprises work together with enterprise methods, via a number of key benefits:

• Improved search capabilities – Amazon Q index augments the generative AI expertise by offering semantically related enterprise content material throughout linked methods via its distributed vector database, delivering question responses at enterprise scale. Now, along with AI for Work, your staff can search information from over 90 connectors, integrating with enterprise methods like Microsoft 365, Salesforce, and Workday whereas additionally connecting with customized inside data methods and third-party search suppliers. AI for Work’s orchestrator manages complicated question processing and agent routing throughout a number of information sources, leading to contextually acceptable and actionable outcomes that considerably cut back search time whereas additionally enabling clever automations that reach far past conventional search capabilities.
• Enhanced information processing – The system constantly ingests and analyzes information via the doc processing pipeline in Amazon Q index, which routinely handles a number of codecs utilizing clever chunking algorithms that protect semantic context. The AI for Work platform unifies search, content material era, and actions in a single interface, to help the creation of multi-step agentic experiences grounded in search. Via real-time incremental indexing that processes solely modified content material, the system maintains information freshness whereas changing siloed uncooked information into actionable insights and multi-step enterprise processes that may be saved and reused throughout the group.
• Value optimization – Organizations can obtain vital value financial savings by streamlining routine duties via brokers that cut back operational overhead and enhance useful resource allocation. AI for Work helps a variety of agent-building choices, from no-code and low-code to pro-code, for each non-technical staff and technical consultants to construct brokers for themselves and to share throughout the group, so groups can accomplish extra with present sources and profit from sustained productiveness enhancements.
• Safety advantages – Safety stays paramount, with Amazon Q index implementing vector-level safety via end-to-end encryption utilizing AWS Key Administration Service (AWS KMS) buyer managed keys and document-level entry controls that filter search outcomes based mostly on consumer identification and group membership. The joint resolution implements strong role-based entry management and audit trails. This zero-trust safety method maintains compliance with business requirements whereas offering granular management over delicate enterprise information, ensuring customers solely see info from paperwork they’ve specific permissions to entry whereas sustaining full information sovereignty. With AI for Work’s strong safety and governance instruments enterprises can handle permissions and agent entry, monitor utilization, and implement guardrails for safe, enterprise-wide deployment of AI options at scale.

Resolution overview

The Amazon Q Enterprise information accessor gives a safe interface that integrates Kore.ai’s AI for Work platform with Amazon Q index. The combination delivers a strong resolution that makes use of enterprise information throughout a number of methods to energy clever agentic actions and content material era capabilities that remodel how organizations deal with routine duties and automate complicated processes end-to-end.

When a consumer submits a question via AI for Work, its orchestrator intelligently routes requests between Kore.ai’s native retrievers and Amazon Q index based mostly on predefined routing guidelines and superior intent recognition algorithms. For Amazon Q index requests, the structure implements safe cross-account API calls utilizing OAuth 2.0 tokens that remodel into non permanent AWS credentials, supporting each safety and optimum efficiency whereas sustaining strict entry controls all through your complete system. With AI for Work’s brokers, customers can take comply with up actions, resembling drafting proposals or submitting tickets—straight on high of search outcomes, for end-to-end process completion in a single interface. Customers can even construct personalised workflows of pre-defined steps and execute them from a single question to additional save time.

This helps use circumstances resembling automated roadmap era, the place a product supervisor can question function requests throughout a number of methods and obtain a structured roadmap full with stakeholder notifications, or RFP response automation, the place gross sales executives can generate complete proposals by pulling compliance documentation and tailoring responses based mostly on shopper necessities.

The next diagram illustrates the answer structure.

Conditions

Earlier than enabling the Amazon Q index integration with Kore.ai’s AI for Work, you will need to have the next elements in place:

• An AWS account with acceptable service entry
• Amazon Q Enterprise arrange with AWS IAM Identification Middle for consumer authentication
• Entry to Kore.ai’s AI for Work (as a workspace admin)

With these conditions met, you may full the fundamental configuration steps on each the Amazon Q Enterprise and Kore.ai consoles to get began.

Add Kore.ai as an information accessor

After creating an Amazon Q Enterprise software with AWS IAM Identification Middle, directors can configure Kore.ai as an information accessor via the Amazon Q Enterprise console. Full the next steps:

1. On the Amazon Q Enterprise console, select Information accessors within the navigation pane.
2. Select Add information accessor.
3. Select Kore.ai as your information accessor. You should retrieve tenantID, a novel identifier in your software tenant. Confer with Conditions for directions to retrieve the TenantId in your software. Related directions are additionally listed later on this put up.
4. For Information supply entry, configure your degree of entry. You possibly can choose particular information sources out of your Amazon Q index to be obtainable via the information accessor. This makes it doable to regulate which content material is surfaced within the AI for Work atmosphere.
5. For Person entry, specify which customers or teams can entry the Amazon Q index via the information accessor. This selection makes it doable to configure granular permissions for information accessor accessibility and handle organizational entry controls.

After you might have added the information accessor, the Amazon Q Enterprise console shows configuration particulars that you should share with Kore.ai to finish the setup.

6. Observe down the next info for the subsequent step:
   1. Amazon Q Enterprise software ID
   2. AWS Area of the Amazon Q Enterprise software
   3. Amazon Q Enterprise retriever ID
   4. Area for IAM Identification Middle occasion

Configure Amazon Q index in Kore.ai’s AI for Work

Kore.ai’s AI for Work helps versatile integration with Amazon Q index based mostly in your enterprise search wants. There are two configuration choices: configuring Amazon Q index as the first enterprise data supply or configuring it as a search agent. We offer directions for each choices on this put up.

Possibility 1: Configure Amazon Q index as the first enterprise data supply

In order for you Amazon Q index to behave as the first fallback search layer, coming into play, full the next steps:

1. In AI for Work, go to Workspaces on the admin console. Then navigate to Enterprise Workspace, which is the default workspace.

2. Select Configure to configure an enterprise data information supply.
3. On the Create New dropdown menu, select Amazon Q.

4. Enter a supply title and transient description.
5. Copy the tenant ID displayed—that is required in the course of the setup of the information accessor in AWS, as described within the earlier part.
6. Enter the small print captured earlier:
   1. Amazon Q Enterprise software ID
   2. Area of the Amazon Q Enterprise software
   3. Amazon Q Enterprise retriever ID
   4. Area for IAM Identification Middle occasion
7. Select Proceed to save lots of and full the configuration.

The brand new data supply now reveals as Lively.

Possibility 2: Configure Amazon Q index as a search agent

If you have already got a main search index, you may configure Amazon Q index as a search agent:

1. In AI for Work, go to Workspaces on the admin console.
2. Select the workspace the place you need to add Amazon Q index. (Enterprise Workspace is utilized by default).
3. Beneath AI Brokers within the navigation pane, select Search Agent
4. Select Create agent.

5. Present an agent title and function. This helps outline when the search agent must be invoked.
6. Select Proceed to maneuver to configuration.
7. For Choose Search Index, select Amazon Q.

8. Copy the tenant ID displayed—it’s required in the course of the setup of the information accessor in AWS.

9. Preview and take a look at the agent.
10. After you might have validated the agent, publish it to chose customers or teams.

Your integration is now full. Now you can entry the assistant software and begin asking questions within the AI for Work console. For those who’ve created a search agent, you may as well entry it from the listing of brokers and begin interacting with it straight.

Clear up

When you’re completed utilizing this resolution, clear up your sources to keep away from extra prices:

1. Disable the Amazon Q index configuration inside AI for Work’s settings.
2. Delete the Kore.ai information accessor from the Amazon Q Enterprise console, which can take away permissions and entry for customers.
3. Delete the Amazon Q Enterprise software to take away the related index and information supply connectors, in your AWS account.

Conclusion

The mix of Kore.ai’s AI for Work and Amazon Q index gives enterprises a transformative method to spice up worker productiveness leveraging complete search capabilities whereas streamlining repetitive duties and processes. By integrating Kore.ai’s superior agentic platform with the strong search infrastructure of Amazon Q index, organizations can now execute context conscious actions by accessing related info throughout disparate methods whereas sustaining information possession and safety. This helps quicker problem-solving, enhanced productiveness, and higher collaboration throughout the group.

On this put up, we explored how enterprises can use the mixing between Kore.ai’s AI for Work and Amazon Q Enterprise to streamline their operational processes and unlock beneficial productiveness features. We demonstrated how organizations can arrange this integration utilizing an Amazon Q information accessor, serving to groups entry essential info securely and cost-effectively.

Unlock the total potential of your group’s information and agentic workflows right this moment with the Amazon Q index and Kore.ai’s AI for Work’s unified resolution by following the steps in Amazon Q integration with AI for Work.

---

In regards to the authors

Siddhant Gupta is a Software program Improvement Supervisor on the Amazon Q group based mostly in Seattle, WA. He’s driving innovation and improvement in cutting-edge AI-powered options.

Chinmayee Rane is a Generative AI Specialist Options Architect at AWS, with a core concentrate on generative AI. She helps ISVs speed up the adoption of generative AI by designing scalable and impactful options. With a robust background in utilized arithmetic and machine studying, she makes a speciality of clever doc processing and AI-driven innovation. Outdoors of labor, she enjoys salsa and bachata dancing.

Bobby Williams is a Senior Options Architect at AWS. He has many years of expertise designing, constructing, and supporting enterprise software program options that scale globally. He works on options throughout business verticals and horizontals and is pushed to create a pleasant expertise for each buyer.

Santhosh Urukonda is a Senior PACE (Prototyping & Cloud Engineering) Architect at AWSs with twenty years of expertise. He makes a speciality of serving to prospects develop progressive, first-to-market options with a concentrate on generative AI.

Nikhil Kumar Goddeti is a Cloud Help Engineer II at AWS. He makes a speciality of AWS Information Analytics providers with emphasis on Amazon OpenSearch Service, Amazon Q Enterprise, Amazon Kinesis, Amazon MSK, Amazon AppFlow, and Amazon Kendra. He’s a Topic Matter Skilled of OpenSearch. Outdoors of labor, he enjoys travelling together with his mates and enjoying cricket.

Meghana Chintalapudi is a Product Supervisor at Kore.ai, driving the event of search and agentic AI options for the AI for Work platform. She has led large-scale AI implementations for Fortune 500 purchasers, evolving from deterministic NLP and intent-detection fashions to superior massive language mannequin deployments, with a robust emphasis on enterprise-grade safety and scalability. Outdoors of labor, Meghana is a dancer and takes motion workshops in Hyderabad, India.

Surabhi Sankhla is a VP of Product at Kore.ai, the place she leads the AI for Work platform to assist enterprises increase worker productiveness. With over 13 years of expertise in product administration and know-how, she has launched AI merchandise from the bottom up and scaled them to tens of millions of customers. At Kore.ai, she drives product technique, shopper implementations, and go-to-market execution in partnership with cross-functional groups. Primarily based in San Francisco, Surabhi is obsessed with making AI accessible and impactful for all.